Inception-Based Network and Multi-Spectrogram Ensemble Applied For Predicting Respiratory Anomalies and Lung Diseases
This work provides a competitive method for detecting respiratory anomalies and lung diseases, which could aid medical professionals in diagnosis.
This paper proposes an inception-based deep neural network to detect lung diseases from respiratory sound inputs. The method achieves competitive ICBHI scores of 0.53/0.45 for respiratory anomaly detection and 0.87/0.85 for disease detection on the ICBHI benchmark dataset.
This paper presents an inception-based deep neural network for detecting lung diseases using respiratory sound input. Recordings of respiratory sound collected from patients are firstly transformed into spectrograms where both spectral and temporal information are well presented, referred to as front-end feature extraction. These spectrograms are then fed into the proposed network, referred to as back-end classification, for detecting whether patients suffer from lung-relevant diseases. Our experiments, conducted over the ICBHI benchmark meta-dataset of respiratory sound, achieve competitive ICBHI scores of 0.53/0.45 and 0.87/0.85 regarding respiratory anomaly and disease detection, respectively.